Participation in substance abuse treatment has consistently proven effective in reducing substance use. However, treatment gains are often not maintained; relapse may occur rapidly after treatment ends and relapse risk remains high for several years. Affiliation with self-help groups (SH) been shown to extend treatment effects. Treatment and SH study findings bear on predictors of short-term abstinence; it is not known whether these findings can be generalized to long-term abstinence. Moreover, SH affiliation studies have typically used treatment samples of alcoholics (often-white men) in Alcoholics Anonymous (AA). Little is known about the effects of SH affiliation among other groups of individuals, especially illicit polysubstance users who may attend Narcotics Anonymous rather than or in addition to AA. Illicit drug users differ from alcohol-dependent persons in critical ways that bear on long-term prognosis and the NA membership differs from AA. Even less is known about the change processes underlying the benefits of SH. The investigation of predictors of long-term abstinence will contribute to enhancing the likelihood of breaking the relapsing cycle of addiction by informing treatment, especially relapse prevention programs, and a longitudinal study of SH affiliation and its underlying processes will inform referral patterns for drug users, elucidate the critical ingredients of affiliation driving therapeutic benefits of SH, a widely used recovery resource, and provide valuable information about the dynamic process of change implicated in the resolution of addiction. This application proposes a longitudinal community-based naturalistic investigation of the predictors of long-term abstinence from illicit drugs with special focus on SH affiliation and underlying change processes. The sample will be a cohort of abstinent individuals (N = 300) stratified by length of abstinence at recruitment (3- to 18-month and 18- to 36months); subjects will be followed-up for 3 years. The aims are: 1. To determine whether predictors of short-term abstinence can be generalized to long-term abstinence. 2. To study longitudinal patterns of affiliation with self-help among illicit drug users. 3. To examine the relationship between these patterns and the maintenance of long-term abstinence. [Aims 2 and 3 would constitute the 1st. long-term, longitudinal investigation of self-help among illicit drug users not recruited in a treatment setting]. 4. To elucidate the processes underlying the demonstrated benefits of affiliation with self-help among drug user by testing a conceptual model (Figure 1) building on and integrating previous findings.